To characterize behavioral deficits in pre-adolescent offspring exposed in utero to Benzo(a)pyrene [B(a)P], timed-pregnant Long Evans Hooded rats were treated with B(a)P (150, 300, 600, and 1200 µg/kg BW) or peanut oil (vehicle) on E14, 15, 16, and 17. Following birth, during the pre-weaning period, B(a)P metabolites were examined in plasma and whole brain or cerebral cortex from exposed and control offspring. Tissue concentrations of B(a)P metabolites were (1) dose-dependent and (2) followed a time-dependence for elimination with ∼60% reduction by PND5 in the 1200 µg/kg BW experimental group. Spatial discrimination-reversal learning was utilized to evaluate potential behavioral neurotoxicity in P40-P60 offspring. Late-adolescent offspring exposed in utero to 600 and 1200 µg/kg BW were indistinguishable from their control counterparts for ability to acquire an original discrimination (OD) and reach criterion. However, a dose-dependent effect of in utero B(a)P-exposure was evident upon a discrimination reversal as exposed offspring perseverated on the previously correct response. This newly characterized behavioral deficit phenotype for the first reversal was not apparent in either the (1) OD or (2) subsequent reversal sessions relative to the respective control offspring. Furthermore, the expression of activity related-cytoskeletal-associated protein (Arc), an experience-dependent cortical protein marker known to be up-regulated in response to acquisition of a novel behavior, was greater in B(a)P-exposed offspring included in the spatial discrimination cohort versus home cage controls. Collectively, these findings support the hypothesis that in utero exposure to B(a)P during critical windows of development representing peak periods of neurogenesis results in behavioral deficits in later life.